From a9e3760b0838299649c0d57cca44daaf40ba3c33 Mon Sep 17 00:00:00 2001 From: Tejun Heo Date: Mon, 14 Sep 2026 22:12:34 -1000 Subject: sched_ext: Maintain an online cid mask in the scheduler arena Schedulers on the default cid mapping treat [0, nr_online_cids) as the online set and restart on hotplug. Schedulers that install their own mapping with scx_bpf_cid_override() have no way to learn which cids are online: the count no longer identifies members and the CPU-form cpumask is unusable from cid programs. This is an obvious hole in the cid API. Add scx_bpf_online_cmask(), a kernel-maintained cmask in the scheduler's arena, allocated alongside the per-CPU scratch masks and populated after the cid mapping is finalized and before ops.init(), for child schedulers too. The pointer stays valid through ops.exit() with no reference to take. It is the arena offset as a void pointer, the same form struct_ops arena arguments arrive in. The verifier types the void return as a scalar for the program's arena cast. The mask follows the SCX hotplug notifications: seeded from cpu_active_mask and updated before ops.cid_online/offline() runs, so it lags cpu_online_mask only inside a hotplug transition. Updates walk the scheduler list under the lock that also serializes unlinking. Reads are live, not atomic snapshots. Root initialization excludes hotplug. v2: Reworded the getter kerneldoc (Andrea Righi). Signed-off-by: Tejun Heo Reviewed-by: Andrea Righi --- include/trace/misc/rdma.h | 168 ++++++++++++++++++++++++++++++++++++++++++++++ 1 file changed, 168 insertions(+) create mode 100644 include/trace/misc/rdma.h (limited to 'include/trace/misc/rdma.h') diff --git a/include/trace/misc/rdma.h b/include/trace/misc/rdma.h new file mode 100644 index 000000000..81bb454fc --- /dev/null +++ b/include/trace/misc/rdma.h @@ -0,0 +1,168 @@ +/* SPDX-License-Identifier: GPL-2.0 */ +/* + * Copyright (c) 2017 Oracle. All rights reserved. + */ + +/* + * enum ib_event_type, from include/rdma/ib_verbs.h + */ +#define IB_EVENT_LIST \ + ib_event(CQ_ERR) \ + ib_event(QP_FATAL) \ + ib_event(QP_REQ_ERR) \ + ib_event(QP_ACCESS_ERR) \ + ib_event(COMM_EST) \ + ib_event(SQ_DRAINED) \ + ib_event(PATH_MIG) \ + ib_event(PATH_MIG_ERR) \ + ib_event(DEVICE_FATAL) \ + ib_event(PORT_ACTIVE) \ + ib_event(PORT_ERR) \ + ib_event(LID_CHANGE) \ + ib_event(PKEY_CHANGE) \ + ib_event(SM_CHANGE) \ + ib_event(SRQ_ERR) \ + ib_event(SRQ_LIMIT_REACHED) \ + ib_event(QP_LAST_WQE_REACHED) \ + ib_event(CLIENT_REREGISTER) \ + ib_event(GID_CHANGE) \ + ib_event_end(WQ_FATAL) + +#undef ib_event +#undef ib_event_end + +#define ib_event(x) TRACE_DEFINE_ENUM(IB_EVENT_##x); +#define ib_event_end(x) TRACE_DEFINE_ENUM(IB_EVENT_##x); + +IB_EVENT_LIST + +#undef ib_event +#undef ib_event_end + +#define ib_event(x) { IB_EVENT_##x, #x }, +#define ib_event_end(x) { IB_EVENT_##x, #x } + +#define rdma_show_ib_event(x) \ + __print_symbolic(x, IB_EVENT_LIST) + +/* + * enum ib_wc_status type, from include/rdma/ib_verbs.h + */ +#define IB_WC_STATUS_LIST \ + ib_wc_status(SUCCESS) \ + ib_wc_status(LOC_LEN_ERR) \ + ib_wc_status(LOC_QP_OP_ERR) \ + ib_wc_status(LOC_EEC_OP_ERR) \ + ib_wc_status(LOC_PROT_ERR) \ + ib_wc_status(WR_FLUSH_ERR) \ + ib_wc_status(MW_BIND_ERR) \ + ib_wc_status(BAD_RESP_ERR) \ + ib_wc_status(LOC_ACCESS_ERR) \ + ib_wc_status(REM_INV_REQ_ERR) \ + ib_wc_status(REM_ACCESS_ERR) \ + ib_wc_status(REM_OP_ERR) \ + ib_wc_status(RETRY_EXC_ERR) \ + ib_wc_status(RNR_RETRY_EXC_ERR) \ + ib_wc_status(LOC_RDD_VIOL_ERR) \ + ib_wc_status(REM_INV_RD_REQ_ERR) \ + ib_wc_status(REM_ABORT_ERR) \ + ib_wc_status(INV_EECN_ERR) \ + ib_wc_status(INV_EEC_STATE_ERR) \ + ib_wc_status(FATAL_ERR) \ + ib_wc_status(RESP_TIMEOUT_ERR) \ + ib_wc_status_end(GENERAL_ERR) + +#undef ib_wc_status +#undef ib_wc_status_end + +#define ib_wc_status(x) TRACE_DEFINE_ENUM(IB_WC_##x); +#define ib_wc_status_end(x) TRACE_DEFINE_ENUM(IB_WC_##x); + +IB_WC_STATUS_LIST + +#undef ib_wc_status +#undef ib_wc_status_end + +#define ib_wc_status(x) { IB_WC_##x, #x }, +#define ib_wc_status_end(x) { IB_WC_##x, #x } + +#define rdma_show_wc_status(x) \ + __print_symbolic(x, IB_WC_STATUS_LIST) + +/* + * enum ib_cm_event_type, from include/rdma/ib_cm.h + */ +#define IB_CM_EVENT_LIST \ + ib_cm_event(REQ_ERROR) \ + ib_cm_event(REQ_RECEIVED) \ + ib_cm_event(REP_ERROR) \ + ib_cm_event(REP_RECEIVED) \ + ib_cm_event(RTU_RECEIVED) \ + ib_cm_event(USER_ESTABLISHED) \ + ib_cm_event(DREQ_ERROR) \ + ib_cm_event(DREQ_RECEIVED) \ + ib_cm_event(DREP_RECEIVED) \ + ib_cm_event(TIMEWAIT_EXIT) \ + ib_cm_event(MRA_RECEIVED) \ + ib_cm_event(REJ_RECEIVED) \ + ib_cm_event(LAP_ERROR) \ + ib_cm_event(LAP_RECEIVED) \ + ib_cm_event(APR_RECEIVED) \ + ib_cm_event(SIDR_REQ_ERROR) \ + ib_cm_event(SIDR_REQ_RECEIVED) \ + ib_cm_event_end(SIDR_REP_RECEIVED) + +#undef ib_cm_event +#undef ib_cm_event_end + +#define ib_cm_event(x) TRACE_DEFINE_ENUM(IB_CM_##x); +#define ib_cm_event_end(x) TRACE_DEFINE_ENUM(IB_CM_##x); + +IB_CM_EVENT_LIST + +#undef ib_cm_event +#undef ib_cm_event_end + +#define ib_cm_event(x) { IB_CM_##x, #x }, +#define ib_cm_event_end(x) { IB_CM_##x, #x } + +#define rdma_show_ib_cm_event(x) \ + __print_symbolic(x, IB_CM_EVENT_LIST) + +/* + * enum rdma_cm_event_type, from include/rdma/rdma_cm.h + */ +#define RDMA_CM_EVENT_LIST \ + rdma_cm_event(ADDR_RESOLVED) \ + rdma_cm_event(ADDR_ERROR) \ + rdma_cm_event(ROUTE_RESOLVED) \ + rdma_cm_event(ROUTE_ERROR) \ + rdma_cm_event(CONNECT_REQUEST) \ + rdma_cm_event(CONNECT_RESPONSE) \ + rdma_cm_event(CONNECT_ERROR) \ + rdma_cm_event(UNREACHABLE) \ + rdma_cm_event(REJECTED) \ + rdma_cm_event(ESTABLISHED) \ + rdma_cm_event(DISCONNECTED) \ + rdma_cm_event(DEVICE_REMOVAL) \ + rdma_cm_event(MULTICAST_JOIN) \ + rdma_cm_event(MULTICAST_ERROR) \ + rdma_cm_event(ADDR_CHANGE) \ + rdma_cm_event_end(TIMEWAIT_EXIT) + +#undef rdma_cm_event +#undef rdma_cm_event_end + +#define rdma_cm_event(x) TRACE_DEFINE_ENUM(RDMA_CM_EVENT_##x); +#define rdma_cm_event_end(x) TRACE_DEFINE_ENUM(RDMA_CM_EVENT_##x); + +RDMA_CM_EVENT_LIST + +#undef rdma_cm_event +#undef rdma_cm_event_end + +#define rdma_cm_event(x) { RDMA_CM_EVENT_##x, #x }, +#define rdma_cm_event_end(x) { RDMA_CM_EVENT_##x, #x } + +#define rdma_show_cm_event(x) \ + __print_symbolic(x, RDMA_CM_EVENT_LIST) -- cgit v1.3.1